Jan 22, 2014

    Bubble trouble here? It's a numbers game

    A BUBBLE can provoke fear or outrage.

    Fear, because if a speculative bubble rising on hot air bursts, the resulting plunge in asset values can bode ill for the economy and cause unemployment and personal hardship. Outrage, because the continued existence of such a bubble implies the failure of regulators to protect the man on the street.

    To say a bubble exists here means there is a lack of control by the powers that be, which is anathema to the Government's perception of itself as a forward-looking architect of a well-planned and orderly economy.

    So one can very well imagine the reaction of Singapore's central bank when Mr Jesse Colombo, a French-born, American independent economic analyst who just turned 28, wrote a widely circulated column on Jan 13 warning that there is not just a low interest rate-fuelled credit bubble here, but also a residential-property bubble, construction bubble, financial-sector bubble, wealth bubble and population bubble.

    The Monetary Authority of Singapore's (MAS') response was swift.

    "Singapore is not facing a credit bubble that puts the country or its banking system at any risk of crisis," the central bank said on the evening of Jan 14.

    While MAS said that unusually low interest rates have caused credit growth and a rise in property prices in Singapore and other countries, the Government has "taken decisive steps to cool property demand and prevent excessive leverage".

    Mr Colombo replied that the MAS was in denial, just like other governments in the region.

    What is a bubble? To decide whether Singapore is indeed caught up in a bubble, it is important, first of all, to define what one is.

    In a bubble, prices of assets deviate significantly from their fundamental values. What constitutes "significant" is arguable. Given sufficient historical data, a possible starting point for a bubble call is when financial ratios deviate from their long-term averages by, say, more than two standard deviations.

    A recent instance of a bubble was the dot-com boom of the late 1990s. Then, the long-term cyclically adjusted price-to-earnings ratios of America's S&P 500 index went up to more than 40 times, more than three standard deviations from the 130-year average of 16.5 times.

    In Singapore, the Straits Times Index is trading at roughly or just under its long-term average of about 15 times earnings. Using our definition, we can safely say there is no bubble in our equity markets.

    Moving on to property, a common indicator of value is the ratio of median house prices to median household yearly incomes. By various accounts, this ratio peaked in America in 2006, almost hitting five times and deviating more than two standard deviations from its long-term average of about 3.5 times. Median household incomes were around US$50,000 then, so median house prices were about US$250,000.

    Here, in the Housing Board (HDB) resale market and the private-property market, average prices have doubled from a decade ago. Incomes have increased by only about 60 to 70 per cent.

    Mr Colombo cited a statistic from cost-of-living comparison website Numbeo that showed Singapore's price-to-income ratio at 25 times - the third-highest globally. On the surface, the number makes sense: The typical price of a suburban condominium is S$1.2 million, while the implied median household income is S$48,000.

    These statistics are flawed, however. The latest household-income data show that households earning S$48,000 a year belong to the bottom third of income earners, who live in much cheaper, subsidised HDB flats anyway.

    More than 80 per cent of Singaporeans live in HDB flats. This puts those who live in condos among the top fifth of wage earners. One cannot benchmark a typical Singaporean to a condo unit.

    There should thus be at least two separate price-to-income ratios for Singapore when measuring whether the market is in a bubble: one for the private-property market, and others for segments of the HDB market.

    Assuming the top fifth of households buy condos, the next fifth buy five-room HDB flats, and so on, back-of-the-envelope calculations using the latest data put the price-to-income ratio for HDB five-room flats at about 4.2 times, and 4.3 times for condos - both far from the statistic of 25 times used in Mr Colombo's piece.

    By comparison, during the property-market trough 10 years ago, the equivalent ratios would be about 3.3 times for five-room flats and 3.6 times for condos.

    Basically, between 2003 and now, homes cost about 0.6 to one year's worth of household income more at the most, from trough to peak. This means people have to take two to three more years to repay their loans, assuming they use a third of their annual incomes to do so. If they use only a fifth of their incomes, then they will take three to five years more.

    If Singaporeans buy property within their means, three or four extra years of paying the mortgage, given later retirement ages and longer lives, seems to be reasonable.

    Interesting tidbits can be gleaned from the analysis. If people aim beyond their wage bracket, house prices become pricey. A typical median-income household buying a five-room resale flat, or even an 81st to 90th percentile household buying a suburban condo, will see their price-to-income ratio shoot up from three-plus to six times - which is arguably unsustainable.

    It is therefore not surprising that, with increased expectations, even the middle- and upper-middle-income groups would complain that houses have become unaffordable. Going by the data, the average Singaporean household can aim for a S$500,000, four-room resale flat to keep the affordability ratio within four times annual income.

    Mr Colombo is not wrong in warning Singapore of the dangers of rising asset prices, if they are not backed by productivity gains or economic growth.

    However, the situation confronting Singapore's residential-property market now is more inflated expectations and less runaway prices.

    Commentators pointing out numerical increases in any statistic should define what makes something a bubble and what does not. Overusing the word can obfuscate rather than illuminate.